Building and Scaling a Product Organization

Adding PMs reactively, as work volume grows, without a deliberate model for dividing ownership, tends to preserve an existing PM's overall scope while adding new PMs alongside them for entirely new initiatives — rather than splitting the original PM's now-overgrown scope into more coherent, individually ownable pieces. This produces organizations where the newest, smallest initiatives are well-owned while the most established, often most important, product areas are owned by a single PM whose attention is spread across far more than they can meaningfully track.

This lesson introduces the Coherence Span Model, plotting a PM's ownership scope by initiative count against interdependency complexity:

A PM's span of coherence — the point past which additional initiatives or complexity make genuinely good, well-informed decision-making impossible — is not a fixed number, since it depends heavily on how interdependent the owned initiatives are; a few tightly related initiatives may be easier to own coherently than the same number of entirely unrelated ones. The Coherence Span Model's discipline is watching for specific signals — decisions increasingly deferred or delayed, decreasing depth of engagement with any single initiative, rising Friction Ledger-style complaints from teams feeling under-supported — that indicate a PM has moved past their sustainable span, rather than waiting for an obvious crisis to force the issue.

A specific organizational mistake is responding to an overextended PM by adding a supporting analyst or associate PM role reporting to them, rather than genuinely splitting ownership into separate, independently-accountable areas. Support roles can help with execution capacity, but they do not solve the underlying coherence problem, since the original PM remains the single point of prioritization and decision-making judgment across an unchanged, overextended scope.

If overextension is this recognizable in hindsight, it's worth asking why it happens so consistently across so many growing companies. Part of the answer is structural: splitting a well-performing PM's ownership can feel, to both the PM and to leadership, like a demotion or a vote of no confidence, even when it is exactly the opposite — a recognition that the area has grown important enough to deserve dedicated, undivided attention. This makes leadership reluctant to initiate the conversation, and PMs reluctant to request it themselves, even once they privately recognize they're stretched thin. The other part of the answer is that overextension develops gradually, one individually reasonable addition at a time, so there is rarely a single obvious moment that triggers a structural review — unlike a hiring decision, which has a clear trigger (new work exists, nobody owns it), a splitting decision requires someone to proactively notice a slow accumulation and decide to act on it before a crisis forces the issue. Naming this reluctance explicitly, and building a periodic, scheduled review that doesn't depend on someone first working up the nerve to raise it, is what allows companies to catch coherence overextension while it's still a manageable adjustment rather than an acute organizational crisis.